5.

THE STABILITY OF A FLOATING BODY

11l1roduct;on

When designing a vessel such as a ship, which is to float on water, it is clearly

necessary to be able to establish beforehand that it will float upright in stable
equilibrium.

Fig 5.1 (a) shows such a floating body, which is in equilibrium under the action of

two equal and opposite forces, namely, its weight W acting vertically downwards
through its centre of gravity G, and the buoyancy force, of equal magnitude W, acting

vertically upwards at the centre of buoyancy B. This centre of buoyancy is located at

the centre of gravity of the fluid displaced by the vessel. When in equilibrium, the
points G and B lie in the same vertical line. At first sight, it may appear that the

condition for stable equilibrium would be that G should lie below 8. However, this is
not so.

I'

~G

w
(a)

(b) Stable

(c) Unslable

Fig 5. J Forces acting on "jloating body

To establish the true condition for stability, consider a small angular displacement
from the equilibrium position, as shown in Figs 5(b) and 5(c). As the vessel tilts, the
centre of buoyancy moves sideways, remaining always at the centre of gravity of the
displaced liquid.

If, as shown on Fig 5(b), the weight and the buoyancy forces

together produce a couple which acts to restore the vessel to its initial position,
the equilibrium is stable. If however, the couple acts to move the vessel even
further from its initial position, as in Fig S(c), then the equilibrium is unstable.
The special case when the resulting couple is zero represents the condition of neutral

stability. It will be seen from Fig 5.1 (b) that it is perfectly possible to obtain stable
equilibrium when the centre of gravity G is located above the centrc of buoyancy B.
In thc following text, we shall show how the stability may he investigated
experimentally, and then how a theoretical calculation can be used to predict the
results.

Experimental Determination ofStability

r= ex
w

._--x. '",""""'/
~--{

'
,!
i-.-._./.
x '."f

l:
UX

'x.--.J

W=wV

(b)

(a)

(c)

Fig 5.2 Derivation ofconditions for stability

Fig 5.2(a) shows a body of total weight W floating on even keel.

The centre of

gravity G may be shifted sideways by moving a jockey of weight \Vj across the width
of the body. When the jockey is moved a distance xi> as shown in Fig 5.2(b), the
centre of gravity of the whole assembly moves to G'. The distance GO', denoted by
xg , is given from elementary statics as

xg

WX
W

_J_J

(5.1)

I' 1

1
1

The shift of the centre of gravity causes the body to tilt to a new equilibrium position.
al a small angle

[0

the vertical, as shown in Fig 5.2(b), with an associated movement

of the centre of buoyancy from B to B

The point B must lie vertically below G.

since the body is in equilibrium in the tilted position.

let the vertical line of the

upthrust through B' intersect the original line of upthrust SG at tbe point M. called the
metacentre. We may now regard the jockey movement as having caused the floating
body to swing about the point M.

Accordingly, the equilibrium is stable if the

mctacentre lies above G. Provided that e is small, the distance GM is given by

Xg

GM~

where

e is

in circular measure. Substituting for xg from Equation (5.1) gives the

result

W. x
GM = - ' . - '
W S

(5.2)
The dimension GM is called the metacentric height.

In the experiment described

below. it is measured directly from the slope of a graph of Xj against

e, obmined by

moving a jockey across a pontoon.

Analy/i:al De/ermina/ion of8M

A quite separate theoretical calculation of the position of the metacentre can be made
as follows.
t

The movement of the centre of buoyancy to B produces a moment of the buoyancy

force about the original centre of buoyancy B. To establish the magnitude of this

1.
l

,,

moment. first consider the element of moment exened by a small clement of change
in displaced volume, as indicated on Fig 5.2(c). An element of width 8x, lying at
distance x from B. has an additional depth

e.x due to the tilt of the body.

Its length.

as shown in the plan view on Fig 5.3(c), is L. So the volume OV cfthe element is

8V

= S.x.L.ox = SLx8x

and the element of additional buoyancy force 8F is

8F

= we Lx8x

w.8V

where \\" is the specific weight of water. The element of momenl about B produced by
the element of force is 8M. where

oM

::: w8Lx 2Ox

= of.x

The total moment about 8 is obtained by integration over the whole of the plan area
of the body. in the plane of the water surface:

M = we fLx 1dX = weI

(5.3)

In this, '1' represents

the second moment, about the axis of symmetry, of the water
,
plane area of the body.
,
t

:'\ow this moment represents the movement of the upthrust wV from B to B namely,
,

wV.BB'. Equating this

[0

the expression for M in Equation (5.3)

wV.BB'

weI

From the geometry of the figure, we see that

BB'

= e.BM

and eliminating BS' between these last two equations gives 8M as

BM

I
V
(5.4)

For the particular case ofa body with a rectangular planfonn of width 0 and length L,
the second moment I is readily found as:
0/1

I;"

0/1 1

fLx ' dX = L fx dx = L

-Dr-

-Df2

[ ']0/1

=
-D/2

LD'
12
(5.5)

42

I_
I_

Now the distance BG may be found from the computed or measured positions of B
th~

and of G, so the metaccntric height GM follows from Equation (S.4) and

lI .,
I .,
I.
I.
lI
III -,III.
II-

geometrical relationship
GM

BM BG

(5.6)
This gives an independent check on the result obtained experimentally by traversing a
jockey weight across the floating body.

Experimental Procedure
The pontoon shown in Fig 5.3 has a rectangular platfonn, and is provided with a rigid
sail. A jockey weight t may be traversed in preset steps and at various heights across
the pontoon, along slots in the sail. Angles of tilt are shown by the movement of a
plumbline over an angular scale. as indicated in Fig 5.3(a).
The height of the centre of gravity of the whole floating assembly is first measured.
for one chosen height of the jockey weight. The pontoon is suspended from a hole at
one side of the sail, as indicated in Fig 5.3(b), and the jockey weight is placed at such
a position on the line of symmetry as to cause the pontoon to hang with its base
roughly vertical. A pumbline is hung from the suspension point. The height of the
centre of gravity G of the whole suspended assembly then lies at the point where the
plumbline intersects the line of symmetry of the pontoon.

This establishes the

position of G for this particular jockey height. The position of G for any other jockey
height may then be calculated from elementary statics, as will be seen later.
After measuring the external width and length of the pontoon. and noting the weights
of the various components. the pontoon is floated in water.
Wilh the jockey weight on the line of symmetry, small magnetic weights are used
trim the assembly to even keel. indicated by a zero reading on the angular scale.

[0

Th~

jockey is then moved in steps across the width of the pontoon. the corresponding

t In some equipmenls. two jockey weights <Ire provided.

experiments, but the <ldaptalions will be obvious.

These gi\'e sLope for slightly diifcr.:nl

angle of tilt (over a range which is typically 8) being recorded at each step. This
procedure is then repeated with the jockey traversed at a number of different heights.

}I'
,

Suspension

Jockey
weight

rl Gi -=

U
-

~-------: ~-~ ~--

--

_/

=- - . --

~-

.~

--

--

Angular
scale
,--- Plumbline

(a) Floating pontoon tilted

(b) Determination ofposition

by movement ofjockey weight

o/centre ofgravity

Fig 5.3 Sketch a/pontoon

Results and Calculations

Weight and Dimensions ofPontoon
Weight of pontoon (excluding jockey weight) W p

2.430 kgf

Weight ofjockey Wj

0.391 kgf

Total weight of floating assembly W = W p + Wj

Iacement V = -W ~ 2.821
P ootoon d ISP
w
1000

2.821 kgf

2.821 x IO-3 m 3

Breadth of pontoon D

201.8 mm = 0.2018 m

Length of pontoon L

360.1 mm

Area of pontoon in plane of water surface

7.267 x 10-2 m 2

LO

0.3601

0.3601 m

0.2018

L0 3
0.360 I x 0.20183
Second Moment of Area I = - - = ---,.,--'---'--"'12
12
3
V
2.821 x 10Depth of immersion OC = - = c:-ccc::--c-,
A
7.267 x 10 2

3.88

10-2 m = 38.8 mm

Height of centre of buoyancy B above 0 08

= BC

OC 19.4 mm

Height o/Centre 0/ Gravity

Fig 5.4 shows schematically the positions of the

centre of buoyancy B. centre of gravity G. and
metacentre M.

0 is a reference point on the

external surface of the pontoon, and C is the point

Y,

the water surface. The thickness of the material

from which the pontoon is made is assumed to be

where the axis of symmetry intersects the plane of

o
Fig 5.4

2 mm. The height of G above rhe reference point

o is OG.
o is Yj'

The height of the jockey weight above

When the pontoon was suspended as shown in Fig

5.3(b) and with the jockey weight placed in the uppennost slot of the sail, the
following measurements were made:
Height ofjockey weight above 0

Yj

Corresponding height of G above 0 OG

345 mm
92 mm

The value of 00 may now be detcnnined for any other value of Yj. If Yj changes by
.6Yj. then this will produce a change in 00 of Wj..6y/W. The vertical separation of
the slots in the sail is 60 mm, so 00 will change in steps 0[0.391 x 60/2.821

Yi

(mm)

105

165

225

285

345

OG

(mm)

58.7

67.1

75.4

83.7

92.0

Table 5.1 Heights DC o/G above base Do/pontoon

8.32

mm. Table 5. J shows the values of 00 calculated in this way for the 5 different
heights Yj of the jockey weight.

45

Experimental determination ofmetaeentric height GM

Table 5.2 shows the re.sults obtained when the pontoon was tilted by traversing the
jockey weight across its width l .
Jockey

Jockey Displacement from Centre,

Xj

(mm)

Height
y; (mm)

-45

-30

-15

15

30

45

105

-7.8

-5.2

-2.7

0.0

2.6

5.2

7.8

165

-6.2

-3.1

0.0

3.2

6.2

225.

-7.7

-3.8

0.0

3.9

7.8

285

-5.2

0.0

5.2

345

-7.5

-0.1

7.4

Table 5.2 Angles oflilt caused by jockey displacement

These results are shown graphically on Fig 5.5. For each of the jockey heights, the
angle of tilt is proportional to the jockey displacement. The metacentric height may
"now be found from Equation (5.2), using the gradients of the lines in Fig 5.5. For
example, when Yi = 105 mm, the gradient is
dX j

5.76 mm/deg

dO

5.76 x 57.3

3330.0mm/rad
mm/rad
330.0

Inserting this into Equation (5.2),

0.391
2.821

GM

330.0

45.7 mm

This value, and corresponding values for other jockey heights, are entered in the
fourth column of Table 5.3. Values of 8M are also shown, derived as follows (refer
to Fig 5.4 for notation):
BM

BG+GM

OG-OB+GM

OG+GM-19.4mm

l The preset sleps in Xj shown in the table are 15 rnm. To provide accuracy, this has been reduced to
7.5 mm in later versions of the equipment.

40
~

-=.-

,, 20
~

Ol--~------~---:::

...-=--~---------I

is.

.'""

>,

~ -20

..,o
u

-40

-8

-6

Fig 5.5 Variation

80

E
E 60

eo

0/ angle of tilt with jockey displacement

BG~BM

------+

~+~

+~

c.:>

0
Angle of tilt

100

-
J

-2

-4

'"

+~
+-

40

20+----'----'----'----'----'-----'
2
3
4
o
1
5
6
Gradient of stability IiDe dx/de (mml")
Fig 5.6 Variation a/stability with me/acentric height

47


Jockey

Metacentric

BM

OG

xj/9

(mm)

(mm/")

105

58.7

5.76

45.7

85.0

165

67.1

4.82

38.3

86.0

225

75.4

3.88

30.8

86.8

285

83.7

2.88

22.9

87.2

345

92.0

2.01

16.0

886

height

.,

height GM

(mm)

(mm)

(mm)

.,

.,

Table 5.3 Me/acentric height derived experimentally

As 8M depends only on the mensuration and total weight of the pontoon, its value

should be independent of the jockey height, and this is seen to be reasonably verified
by the experimental results. The value computed from theory is

8M

1
V

2.466
2.821

X
X

10-4
10-3

8.74

10-2 m

87.4 mm

which is in satisfactory agreement with the values obtained experimentally.

Another way of expressing the experimental results is presented in Fig 5.6, where the
height BG of the centre of gravity above the centre of buoyancy is shown as a

function of the slope

x/e.

The experimental points lie on a straight line which

intersects the BG axis at the value 90 mm. As BG approaches this value, x/S ----). O.
Namely, the pontoon may be then tilted by an infinitesimal movement of the jockey
weight; it is in the condition of neutral stability. Under this condition, the centre of
gravity coincides with the metacentre, viz. BM
BM

90 mm.

.
..

BG. So, from Fig 5.6, we see that

This experimentally detennined value again is in satisfactory

agreement with the theoretical value of 87.4 mm.

..

Di.,.cussiOIl ofResults

The experiment demonstrates how the stability of a floating body is affected by

changing the height of its centre of gravity, and how the metacentric height may be

established experimentally by moving the centre of gravity sideways across the body.
The value established in this way agrees satisfactorily with that given by the
analytical result BM = JlV.